Apple is reportedly ramping up its financial commitment to artificial intelligence (AI) infrastructure in the coming years. According to insights from analyst Ming-Chi Kuo, the tech giant is allocating substantial funds towards its generative AI advancements.
Upcoming iPads Slated for March Release, Suggests Insider Report
For 2023, Apple is projected to channel upwards of $620 million towards specialized AI servers, and this figure is expected to balloon to approximately $4.75 billion by 2024. The company’s buying patterns suggest a potential acquisition of 2,000 to 3,000 servers this year, with this number anticipated to climb to nearly 20,000 by next year.
Kuo’s analysis points towards Apple’s interest in Nvidia’s 8-GPU HGX H100 servers for their generative AI endeavors, with an eye on transitioning to the B100 variant in the subsequent year. Nvidia touts the H100 as a premium AI supercomputing solution, each carrying a price tag in the ballpark of $250,000.
Apple’s inclination seems to be towards utilizing these AI servers for in-house training of extensive language models. This choice, rather than opting for external cloud services, likely stems from a desire to bolster data security and user privacy. There’s also speculation around Apple potentially venturing into crafting its own server chips to drive down associated expenses, although this remains unconfirmed.
Yet, even with these aggressive investment strategies in the realm of AI, Apple might find itself trailing industry leaders like Meta and Microsoft. The entire ecosystem, encompassing labor, infrastructure, and other facets, demands consistent multibillion-dollar annual investments. Kuo expresses concerns regarding Apple’s AI future if their spending hovers around the $1 billion mark annually, as previously hinted by Bloomberg’s Mark Gurman.
